Maybe with a reversed flow : not the one who finds the most, but the one in whom the least is found should be rewarded. I guess we can’t rule out other biases.
The best way I’ve found is by using a combination of measures to counter Goodhart’s Law. I’ve also discussed it at length in the post that’s linked in the article, if you’re interested in checking it out.
“Manipulate” sounds like a controversial word in management. But what you’re describing is probably the most accurate, and actionable description of leadership.
The image of the leader on a white horse, riding ahead with vision and glory is outdated. An effective leader is closer to a craftsperson, shaping the environment, narratives, values with care and intention.
The environment lever is the one I come back to most as an engineering manager.
A few years ago my team was constantly complaining about slow delivery. The usual response would have been a talk about ownership or urgency. Instead I looked at the system.
We had accumulated a series of approval gates code reviews requiring two senior sign-offs, deployment approvals, change request forms each added at some point for a good reason, none of them ever removed. Nobody had noticed they'd turned into a queue. The seniors were the bottleneck, and they didn't even know it.
We audited every gate. Removed the ones that added friction without adding safety. Made automated tests the gate instead of people. Delivery speed improved within weeks not because the team suddenly became more motivated, but because we stopped making good behavior hard.
The line about "people respond to structure far more reliably than speeches" is the most underused insight in engineering leadership. Most managers I've seen reach for the speech first. The structure conversation never happens.
The behavior my system was quietly encouraging before that change was learned helplessness waiting for approval rather than moving forward. The structure was teaching that, not the people.
“Behavior rarely changes because of instructions alone.” That line exposes so much.
We tend to coach effort when we should be examining structure. Environment reveals what we truly value — not what we say we value.
Really thoughtful framing.
Very interesting, thank you.
Despite the side effects, I think running an internal bug bounty program is a really smart move.
Maybe with a reversed flow : not the one who finds the most, but the one in whom the least is found should be rewarded. I guess we can’t rule out other biases.
The best way I’ve found is by using a combination of measures to counter Goodhart’s Law. I’ve also discussed it at length in the post that’s linked in the article, if you’re interested in checking it out.
Agreed - no harm in the program itself. What matters is how the incentives are set and measured.
I’m a firm believer of systems thinking applied to management, but had not thought it through as thoroughly as you did: thank you!
If interested, I wrote something in the same ballpark a while back - you can find it here: https://open.substack.com/pub/theintentionalmanager/p/the-hidden-enemy-of-change-your-systems?r=5bq0ac&utm_medium=ios
Thanks for the share, Luca, will check it out
“Manipulate” sounds like a controversial word in management. But what you’re describing is probably the most accurate, and actionable description of leadership.
The image of the leader on a white horse, riding ahead with vision and glory is outdated. An effective leader is closer to a craftsperson, shaping the environment, narratives, values with care and intention.
The environment lever is the one I come back to most as an engineering manager.
A few years ago my team was constantly complaining about slow delivery. The usual response would have been a talk about ownership or urgency. Instead I looked at the system.
We had accumulated a series of approval gates code reviews requiring two senior sign-offs, deployment approvals, change request forms each added at some point for a good reason, none of them ever removed. Nobody had noticed they'd turned into a queue. The seniors were the bottleneck, and they didn't even know it.
We audited every gate. Removed the ones that added friction without adding safety. Made automated tests the gate instead of people. Delivery speed improved within weeks not because the team suddenly became more motivated, but because we stopped making good behavior hard.
The line about "people respond to structure far more reliably than speeches" is the most underused insight in engineering leadership. Most managers I've seen reach for the speech first. The structure conversation never happens.
The behavior my system was quietly encouraging before that change was learned helplessness waiting for approval rather than moving forward. The structure was teaching that, not the people.
Such a great story and insight, Diamantino, thank you for sharing!